[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

Re: [ba-unrev-talk] Response from ADM to Eric Armstrongs review of our product.

Great response, Eric. You're definitely taking the feedback in the
right vein. I admire your tenancity. It can take you a long way. I
tend to think that the environment was very different for Adope
(Adobe) and Logus (Lotus) -- (I like the typo'd versions so much
better...) -- but then, the people who say you can't *always* have
a reason why it's different now, so please do prove me wrong.    (01)

I'm looking forward to your talk with your CTO. The XML decision
is a no-brainer, but a Google-style back-link monitor is a major
investment, plus a potential performance bottleneck. *Someday*,
we'll clearly be there. But if you go there too soon, you fail, and if
you go too late, you're an also-ran. Timing it right is *really* tough,
and mostly a matter of luck.    (02)

Finally, I have to admit to having a one-track mind with respect to
outliner interfaces. I'm used to collapsible hierarchies and don't
have much "time in grade" with multi-column or other approaches,
so there may well be productivity advantages I have missed.    (03)

eric    (04)

ADM Staff wrote:    (05)

> Hi there,
> Thank you *very*much for the encouraging yet appropriatelty critical
> review of ADM below.
> Your suggestion - and that of Eugene - for incorporating xml into
> version 2.0 will be discussed with my partner Arne, who is also the
> CTO of our company, (he may already have planned for it) when I reach
> Vancouver later this week.
> I will interpolate responses to your other remarks below.
> ========
> By the way, thank you for the demo you gave of your system the other
> day. I have a lengthy review to post one day, but for now I'll try to
> hit the
> highlights.
> ADM is clearly "more than an outliner". It looks like:
> a) A great web browser that promotes "bookmarks" to first-class
> citizens, making them outline-entries that can be easily copied and
> organized.
> ===== Agreed.  Marking and accessing all data types including urls is
> a key element in adm.
> In addition:
> b) It's an authoring tool
> c) Pages can be annotated with additional information.
> d) You can link to other outline structures, which makes easier to
> "graft on" other structures.
> ===== all of above is correct from my perspective.  Moreover, the
> forthcoming addition in of linking to urls, files, and topics in the
> pages (currently only the topics accept links); and the addition of
> multiple windowing, together with ability to emulate webpage type
> styles in the pages, will *greatly* augment the authoring and
> datasharing capabilities in version 2.0
> However, to my mind it falls short of being a "knowledge management
> desktop", because there did not appear to be any good way to import
> the
> *collective* wisdom of the group -- or more importantly, the
> collective wisdom of the
> global web community.
> For example, the ability to annotate pages has tremendous potential
> for
> adding Amazon-style ratings. But unless I can find all such entries,
> made by
> anyone who visited that page (using backlinks), then I can never get a
> complete
> picture of what others are thinking. (Given such capabilities, the
> next step
> would be to merge the! r! ankings, Amazon-style, but a google-style
> backlink-server
> would be a minimum necessary condition for that capability.)
> ====== I will discuss this idea with Arne.
> The system we looked at (and the upcoming version we discussed) had
> two
> other notable limitations, as well: a) A proprietary, binary data
> format for the outline structures. More than anything else, this
> choice will limit wide-spread adoption,
> because the structures will be neither as easily readable or as
> readily
> constructable as HTML or XML documents. (I hope, therefore, that
> you are considering switching to an XML format, in future.)
> ======  Yes, this is an option I will discuss with Arne.
> b) A windows-only implementation.
> A "reference implementation" is needed that the open source folks
> can utilize and extend on a variety of platforms -- because *no* web
> based solution can hope to achieve ubiquity if it only runs on a
> single
> platform.
> (And if multiple platform implementations must depend on a single
> small company acquiring the resources to develop them, then I fear
> that the effort is doomed from the outset.)
> =====  Maybe not doomed.  We are very tenacious.  Also, logus notes
> and adobe are counter-examples of formats which have achieved
> widespread adoption.
> Finally, a word on the interface. I was favorably impressed my much
> that
> I saw, including the various mechanisms for acquiring entries and
> shuffling them around.
> ===== Thank you.  I am personally very proud of the drag-and-drop
> capabilities for linking all file types and urls into adm datat
> structures.  A key upgrade in version 2.0 will be ability to also
> browse *all* file types in the internal adm browser, rather viwing
> only urls in the internal browser and needing to view other file types
> in their native applications.
> Although I've never been a big fan of the Lotus-style 3-column
> hierarchy implemented in the current version, I look forward to the
> collapsible
> tree.
> ====== Slight correction to above:  There are currently 5 different
> dataviews - one, two, three panel, cards right for viewing documents
> and url's, and short-long two panel.  Also, as you note, a collapsable
> outliner will be added.  Moreover, the present `speed screens' have
> there own advantages in terms of *blazingly fast' construction and
> navigation of datastructures, unmatcheable by any tree or collapsable
> view.  This aspect has to be experienced with realworld tasks to
> appreciate just what an advantage it affords.  However, as the
> collapsable view has its own merits, in terms of juxtaposing different
> extended outline topic branches, and in terms of preparing linear
> reports, etc., we are definitely set to add this capability in version
> 2.0.
> Unfortunately, as you are no doubt finding out, collabsible tree
> components only allow for single-line entries, rather than multiple
> lines -- a
> factor which more than any other has limited the functionality of any
> GUI-based
> outlines which have been developed in the last couple of decades.
> ======  I'm not clear on what you mean by the above.   To my
> knowledge, our collabsable view will, like the current views, include
> multiple line viewing of any topic.  Have I missed something here?
> Thanks again for your detailed feedback below.  I intend to share all
> of it with my associates.
> Very Cordially, Eric Sommer, CEO, ADM
> luck.
> eric
> Advanced Data Management Systems
> Email: staff@adm21.net
> WorldWideOutline Initiative:
> Download free Trial Client:
> http://www.worldwideoutline.net
> ADM homepage:
> http://www.adm21.net
> -----------------------------------------------------------------------
> Do you Yahoo!?
> HotJobs - Search new jobs daily now    (06)